AI-ENABLED REGRESSION AND TIME SERIES MODELS FOR FORECASTING DIGITAL ECONOMY TRENDS

Accurate forecasting is critical for Iraq’s digital economy, where fragile infrastructure and weak governance limit resilience. This study examined how regression, time series, and hybrid AI models shaped outcomes in growth prediction accuracy, sector monitoring, employment forecasting, and policy support between 2020 and 2024. A descriptive design using 25 sector-year observations from secondary data guided the analysis. Correlation results showed strong positive links between digital outcomes and hybrid models at 0.81, regression at 0.76, and time series at 0.72, while structural barriers had a negative effect at −0.58. Regression analysis confirmed hybrid models as the strongest driver with a coefficient of 0.37, followed by regression at 0.26 and time series at 0.22, while barriers reduced outcomes at −0.19. The model explained 80 percent of the variance, validating robustness. Findings showed forecast accuracy improved from 55 to 75 percent, forecast error dropped from 45 to 22 percent, GDP forecasts aligned within 2 percent deviation, and adoption in banking and retail rose from under 10 to 20 percent. Employment forecasts improved accuracy from 45 to 65 percent but remained constrained by poor records. These results imply that Iraq’s resilience depends on scaling hybrid models, strengthening infrastructure, and embedding forecasting into governance. Recommendations highlight investment in broadband, cloud, and training, scaling regression and hybrid models in ministries and enterprises, and embedding applied forecasting in higher education.